Optimal control of a Markovian failure-prone manufacturing system under a risk-averse cost criterion

The optimality of a hedging control policy in a Markovian failure-prone manufacturing system subject to a constant rate of demand for parts is established for a long-run risk-averse criterion, which is the conditional value-at-risk of the steady-state instantaneous running cost. This extends the known classical result of optimality of hedging policies in failure prone manufacturing systems under the long-run average cost that is a risk-neutral criterion.